B1.2 Unit Scientific Objectives. While creating new systems and methods that efficiently meet research data needs, the DMIS Unit will also make those new systems and methods an object of study. To the extent that these novel approaches are proven by formal evaluations to be better than traditional data management approaches, we will publish those evaluations as contributions to the peer-reviewed literature. B1.3 Innovations. An important focus of the Data Management Unit's innovation is creating an infrastructure for distance-independent national and international collaboration using the Internet. For example using a database-driven approach to dynamic web publication ('portal'approach), web pages are dynamically constructed via scripts that assemble each page 'on the fly'as users click to request various types of information. The design of our data model for the project emphasizes representational power and flexibility to store variable numbers of time-oriented repeated measures. Toward this goal, we employ both traditional unit record- oriented data modeling and Entity-Attribute-Value (EAV) data tables that permit the creation of new variables and fields simply by defining new entries in a system data dictionary.^'^ This structural flexibility combined with our dynamic web-based suite of utilities provides the framework for rapid deployment of data resources in support of new studies.